Submucosal endoscopy with mucosal resection: A hybrid endoscopic submucosal dissection in the porcine rectum and distal colon

Background: Endoscopic submucosal dissection (ESD) has appeal for en bloc resection of large flat colorectal polyps but carries appreciable risk and demands a high level of expertise and training. Undermining flat and laterally spreading colorectal polyps by using submucosal endoscopy with the mucosal safety valve flap (SEMF) may be combined with ESD to offer a hybrid technique that is easier and safer. Objective: To determine the feasibility of combining SEMF and ESD for the removal of progressively larger areas of the mucosa in the porcine rectum and colon. Design: Two-phase ex vivo and in vivo study using domestic pig rectum and distal colon. Setting: Developmental endoscopy unit/animal research unit. Interventions: Progressively larger targeted mucosal dissections were performed by using the basic principles of ESD, which included margination of the targeted area of mucosa, submucosal fluid cushion, and needle-knife dissection. These were combined with the SEMF method of predissection with carbon dioxide (CO₂) gas and balloon-based blunt dissection of the submucosa undermining the targeted and isolated mucosa. The hybrid technique was first applied to ex vivo porcine rectums and distal colons, then in vivo in an acute animal study. Progressively larger staged dissections were performed with 2-, 4-, and 6-cm diameter targeted mucosal sites. Main Outcome Measurements: Success with associated difficulty or failure of the hybrid method in the rectum and distal colon to achieve complete resection of a progressively larger targeted area of mucosa. Results: The ex vivo phase of the experiment demonstrated the ability to use the blunt balloon dissection of the SEMF procedure to remove 2-, 4-, and 6-cm areas of rectum and distal colon with a rapid progression to the largest size resection. The colon proximal to 20 cm above the anus was unable to hold a submucosal fluid cushion and allow submucosal dissection. Successful hybrid ESD was performed in vivo with staged progression through to the largest mucosal area (6 cm) within 20 cm of the anus. Dissections became progressively easier and faster to perform and dependent on the following steps: a traditional circumferential mucosal incision into the deep submucosa, an initial needle-knife submucosal "tunnel" dissection above the muscularis propria (MP), balloon dissection, and needle-knife release of persistent tethering strands of submucosa with and without vessels. Of 16 hybrid resections, 3 failures and 2 very difficult resections were attributed to attempts at resection too proximal in the colon, excessive flexibility of the balloon catheter, and electrosurgical device subperformance. There were no mucosal perforations or coagulation injuries to the MP. A single uninflated balloon catheter perforation of the MP occurred in a dissection site just above 20 cm from the anus with a suboptimal fluid cushion. Limitations: Animal study. Procedures performed by a single endoscopist with long-standing familiarity with the SEMF method. Conclusions: Large mucosal target sites in the rectum and distal colon of the pig can be safely removed en bloc by means of a hybrid technique, ie, submucosal endoscopy with mucosal resection, combining elements of ESD with our SEMF method.